Hasni Hassan

Solving Incomplete Datasets in Soft Set Using Parity Bits of Supported Sets

The theory of soft set proposed by Molodtsov [2]in 1999 is a new method for handling uncertain data and can be redefined as a Boolean-valued information system. The soft set theory has been applied to data analysis and decision support systems based on large datasets. Using retrieved datasets, we have calculated the supported values and then determine the even parity bits. Using the parity bit, the problem of missing values from the retrieved datasets can be solved.

Hybrid reduction in soft set decision making

In this paper, we present an extended technique of decision making by implementing column reduction with reduction based on calculated maximal support objects. Using a Boolean valued information system, certain rows or objects can be defined as ultimate maximum support object, ultimate minimum support object and zero significance parameter. One can then reduce a table by eliminating the defined row or objects in what has been defined as hybrid reduction. As part of our paper, we have managed to show that our proposed model of hybrid reduction yielded a better data size reduction whilst still maintaining consistent results.

Solving Incomplete Datasets in Soft Set Using Supported Sets and Aggregate Values

Abstract The theory of soft set proposed by Molodtsovin 1999[1]is a new method for handling uncertain data and can be defined as a Boolean-valued information system. Ithas been applied to data analysis and decision support systems based on large datasets. In this paper, it is shown that calculated support value can be used to determine missing attribute value of an object. However, in cases when more than one value is missing, the aggregate values and calculated support values will be used in determining the missing values. By successfully recovering missing attribute values, the integrity of a dataset can still been maintained.

Algorithms for extracting adjacency matrix based on formal concept analysis (FCA)

The desire to achieve a holistic representation of Information Retrieval (IR) with the aim for a human-oriented form of representation has spurred the growth of concept-based IR search techniques such as the Semantic Web technology. However, Semantic Web calls for the use of ontologies for many domains. Although meaningful and important, ontology development presents great challenges to the developers especially in terms of conceptual dynamics.. This paper is based on a study that attempts to provide an alternative to ontology lookup for Semantic information retrieval. However, the focus of the paper is on a method proposed to extract adjacency matrix from concepts obtained from the theory of Formal Concept Analysis (FCA) using two consecutive algorithms called the Relatedness Algorithm and Adjacency Matrix Algorithm. Consequently, the adjacency matrices obtained could be used in a similarity measure process based on graph theory. The proposed method offers an alternative to specific domain ontology look-up where results from the measure can further be used in concept-based IR process.

DQAS implementation to support Base Analysis techniques in Data Quality Life Cycle

Good business decision making depends on how good information is provided. Due to this factor, the quality of data provided by transactions systems database is really important to organizations in order to produce the best solution for their company to move forward. Data Quality issues have become major problems in most enterprise systems where in some cases forced companies to stop their operations and in other, create chaos in life environment. Base Analysis technique is part of analyses which are introduced in our previous research in Data Quality Life Cycle. Base Analysis technique is used to profile heterogeneous data in a structured approach, with the intention to determine abnormal data. The technique contains three levels of analyses; Top Level Analysis, Middle Level Analysis and Low Level Analysis. On the other hand, Data Quality Analysis System is a tool developed using open source technologies that are connected to commercial databases to support Base Analysis and implemented in three-tier architecture. This paper describes the implementation of the system to perform data quality analysis using Human Resources data that reside inOracle database platform.